This is a study of behavioral factors in medical outpatients with mild hypertension. The investigators will examine: (1) the behavioral side effects (memory deficits and depressive symptoms); and (2) the efficacy in diastolic blood pressure (DBP) control of drug and nondrug strategies in the treatment of mild hypertension. The target population will be persons with mild hypertension (DBP - 90-104 mm Hg) between 18 and 59 years old working or living in San Francisco. A total of 500 participants will be recruited over an 18 month accrual period from collaborating health facilities at the UCSF Medical Center and from referrals made by the Bank of America and the Bay Area Hypertension Council screening program. Potential participants will be evaluated for eligibility and exclusion criteria during two clinic visits and baseline psychological measures will be obtained at a third clinic visit. Participants will be randomized using a stratified blocked design to one of four treatment assignments: metoprolol, placebo, nonpharmacologial plus placebo, and nonpharmacological plus metropolol. The nonpharmacological treatment is a behavioral intervention to change dietary habits and other lifestyle factors which contribute to the presence of elevated blood pressure. The pharmacological treatment with a beta adrenergic blocking drug is a standard approach to treating mild hypertension. At least 15 visits will be scheduled over the one year follow-up period and nurse practitioners will be the health care providers. The factorial design will permit the effects of these two interventions to be assessed separately and in combination. The primary outcome measurement will be memory deficits, depressive symptoms and DBP. The psychological variables will be asssessed at 3 and 12 months and DBP will be obtained at each visit. This study will have a power of at least .80 to detect a significant difference (Alpha = 0.05) in memory deficits, depressive symptoms and/or DBP control between the treatment groups. The study will also examine, though with less power, the efficacy of the interventions in specified subgroups of persons with mild hypertension. If the nonpharmacologic intervention is effective in lowering blood pressure with less biobehavioral sequelae, it will be applicable to persons with mild hypertension in other ambulatory care settings of the country. This study will also serve as a model for further investigation of nonbiobehavioral sequelae of other antihypertensive regimes.